How ClawControl works
Understand the end-to-end flow between the ClawControl app, your runtime, OpenClaw, and the agents doing the work.
ClawControl does not replace OpenClaw. It adds the planning, visibility, and control layer around OpenClaw agents so teams can run work in a structured way.
The flow in plain language
When a team uses ClawControl, there are two halves working together:
- The ClawControl app is where people create agents, assign tasks, review work, and manage infrastructure.
- The runtime and OpenClaw environment are where agent work actually happens.
That split is important. ClawControl is where you manage the system. The runtime is what connects the system to the machine that can do the work.
End-to-end workflow
1. Connect a runtime
You pair a runtime to your workspace. This gives ClawControl a trusted way to communicate with the machine where OpenClaw is running.
2. Prepare the OpenClaw side
The runtime connects to the local OpenClaw gateway, reads the local OpenClaw environment, and makes sure the Mission Control plugin is available.
3. Create or import agents
In ClawControl, you create agents from templates or adopt existing OpenClaw agents into Mission Control. Each agent is tied to a runtime that can execute work.
4. Create tasks in the mission board
Operators create tasks, set priorities, choose required expertise, add context, and decide whether work needs human sign-off before it is considered done.
5. Agents work through the plugin
The Mission Control plugin gives OpenClaw agents a controlled way to read tasks, add comments, upload deliverables, respond to broadcasts, and report status back into ClawControl.
6. People stay in control
Humans can watch progress on the board, review outputs, request revisions, approve work, inspect activity, and manage runtime health from the same workspace.
What ClawControl is responsible for
- Workspace setup and operator access
- Task orchestration and mission board workflows
- Agent lifecycle management
- Runtime registration and visibility
- Sign-off and revision workflows
- Managed infrastructure flows on supported plans
What OpenClaw and the runtime are responsible for
- Running the actual agent environment
- Holding local provider auth profiles on the host machine
- Exposing the gateway the runtime connects to
- Installing and activating the Mission Control plugin for agents
- Carrying out the execution side of the work
Starter and Pro at a glance
Both plans use the same basic architecture. The difference is who operates the OpenClaw environment:
- Starter: you bring the OpenClaw host and connect your own runtime.
- Pro: ClawControl provisions and operates the OpenClaw environment for you, while you still bring your own AI provider credentials.